Abstract Bone stress injury (BSI) is commonly seen among highly active athletes due to long and intense activities in sports. Weight-bearing long bones, such as the tibia and fibula, are among the most common sites of BSI in active individuals. Microscopic crack formation, increased load transference to bone, excessive and repetitive mechanical loads, and cyclic loading and contractions on the vascular system the following phenomena all contribute to BSI. All hypothesized mechanisms end with the occurrence of a set of microcracks and collagenous matrix rupture. Early diagnosis is crucial for optimal treatment with the goal of minimizing disability. Clinical MRI has become the gold standard for BSI diagnosis at the early stage. However, clinical MRI focuses on visualization of edema in marrow and the surrounding soft tissues rather than the injured components of bone. In particular, normal and injured bone both have very short T2/T2* and are ?invisible? with conventional clinical MRI. It would be a major achievement to develop MRI techniques to evaluate clinical MRI ?invisible? cortical bone, facilitating BSI diagnosis and treatment monitoring. We have developed Ultrashort Echo Time (UTE) sequences with TEs of 8 s that are 100~1000 times shorter than conventional TEs of several milliseconds. This makes it possible to directly detect multiple water components, including water residing in the macroscopic pores (pore water) and water bound to collagen (bound water). Bound water may be selectively imaged using a double adiabatic inversion recovery prepared UTE (DIR-UTE) sequence. The combination of UTE and magnetization transfer (UTE-MT) imaging and signal modeling provides indirect evaluation of collagen backbone protons, including their fractions and exchange rates with water protons. UTE with quantitative susceptibility mapping (UTE-QSM) can map bone susceptibility, providing indirect measurement of bone mineral content. The goal of this study is to investigate UTE sequences for comprehensive evaluation of bound and pore water as well as collagen and mineral in cortical bone of specimens, healthy volunteers and athletes subject to BSI. To achieve this goal, in Aim 1 we will validate 3D UTE sequences (UTE, IR-UTE, UTE-MT and UTE-QSM) in quantifying bound and pore water as well as collagen and mineral in tibias and fibulas of 40 leg specimens from two groups of donors (<40y, n=20; >70y, n=20), before and after dynamic loading. We hypothesize that UTE MRI techniques can assess all major components (water, collagen, mineral) in cortical bone, and UTE metrics are highly correlated with changes in biomechanical properties and microcracks quantity induced by dynamic loading. In Aim 2 we will apply the 3D UTE sequences to evaluate total, bound, and pore water as well as collagen and mineral in healthy volunteers (n=20) and athletes (n=20) subject to BSI. We hypothesize that UTE measures will better characterize longitudinal bone changes in athletes subject to BSI, improving diagnosis and therapeutic monitoring compared with current gold standard MRI sequences.